Mercurial > pylearn

annotate algorithms/aa.py @ 500:3c60c2db0319

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Added new daa test
author Joseph Turian <turian@gmail.com>
date Tue, 28 Oct 2008 13:36:27 -0400
parents 8fcd0f3d9a17
children
rev   line source
476
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
1
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
2 import theano
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
3 from theano import tensor as T
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
4 from theano.tensor import nnet as NN
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
5 import numpy as N
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
6
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
7 class AutoEncoder(theano.FancyModule):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
8
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
9 def __init__(self, input = None, regularize = True, tie_weights = True):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
10 super(AutoEncoder, self).__init__()
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
11
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
12 # MODEL CONFIGURATION
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
13 self.regularize = regularize
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
14 self.tie_weights = tie_weights
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
15
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
16 # ACQUIRE/MAKE INPUT
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
17 if not input:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
18 input = T.matrix('input')
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
19 self.input = theano.External(input)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
20
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
21 # HYPER-PARAMETERS
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
22 self.lr = theano.Member(T.scalar())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
23
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
24 # PARAMETERS
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
25 self.w1 = theano.Member(T.matrix())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
26 if not tie_weights:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
27 self.w2 = theano.Member(T.matrix())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
28 else:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
29 self.w2 = self.w1.T
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
30 self.b1 = theano.Member(T.vector())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
31 self.b2 = theano.Member(T.vector())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
32
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
33 # HIDDEN LAYER
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
34 self.hidden_activation = T.dot(input, self.w1) + self.b1
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
35 self.hidden = self.build_hidden()
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
36
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
37 # RECONSTRUCTION LAYER
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
38 self.output_activation = T.dot(self.hidden, self.w2) + self.b2
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
39 self.output = self.build_output()
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
40
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
41 # RECONSTRUCTION COST
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
42 self.reconstruction_cost = self.build_reconstruction_cost()
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
43
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
44 # REGULARIZATION COST
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
45 self.regularization = self.build_regularization()
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
46
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
47 # TOTAL COST
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
48 self.cost = self.reconstruction_cost
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
49 if self.regularize:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
50 self.cost = self.cost + self.regularization
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
51
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
52 # GRADIENTS AND UPDATES
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
53 if self.tie_weights:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
54 self.params = self.w1, self.b1, self.b2
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
55 else:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
56 self.params = self.w1, self.w2, self.b1, self.b2
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
57 gradients = T.grad(self.cost, self.params)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
58 updates = dict((p, p - self.lr * g) for p, g in zip(self.params, gradients))
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
59
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
60 # INTERFACE METHODS
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
61 self.update = theano.Method(input, self.cost, updates)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
62 self.reconstruction = theano.Method(input, self.output)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
63 self.representation = theano.Method(input, self.hidden)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
64
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
65 def _instance_initialize(self, obj, input_size = None, hidden_size = None, seed = None, **init):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
66 if (input_size is None) ^ (hidden_size is None):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
67 raise ValueError("Must specify hidden_size and target_size or neither.")
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
68 super(AutoEncoder, self)._instance_initialize(obj, **init)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
69 if seed is not None:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
70 R = N.random.RandomState(seed)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
71 else:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
72 R = N.random
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
73 if input_size is not None:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
74 sz = (input_size, hidden_size)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
75 range = 1/N.sqrt(input_size)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
76 obj.w1 = R.uniform(size = sz, low = -range, high = range)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
77 if not self.tie_weights:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
78 obj.w2 = R.uniform(size = list(reversed(sz)), low = -range, high = range)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
79 obj.b1 = N.zeros(hidden_size)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
80 obj.b2 = N.zeros(input_size)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
81
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
82 def build_regularization(self):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
83 return T.zero() # no regularization!
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
84
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
85
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
86 class SigmoidXEAutoEncoder(AutoEncoder):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
87
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
88 def build_hidden(self):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
89 return NN.sigmoid(self.hidden_activation)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
90
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
91 def build_output(self):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
92 return NN.sigmoid(self.output_activation)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
93
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
94 def build_reconstruction_cost(self):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
95 self.reconstruction_cost_matrix = self.input * T.log(self.output) + (1.0 - self.input) * T.log(1.0 - self.output)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
96 self.reconstruction_costs = -T.sum(self.reconstruction_cost_matrix, axis=1)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
97 return T.sum(self.reconstruction_costs)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
98
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
99 def build_regularization(self):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
100 self.l2_coef = theano.Member(T.scalar())
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
101 if self.tie_weights:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
102 return self.l2_coef * T.sum(self.w1 * self.w1)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
103 else:
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
104 return self.l2_coef * T.sum(self.w1 * self.w1) + T.sum(self.w2 * self.w2)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
105
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
106 def _instance_initialize(self, obj, input_size = None, hidden_size = None, **init):
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
107 init.setdefault('l2_coef', 0)
8fcd0f3d9a17 added a few algorithms
Olivier Breuleux <breuleuo@iro.umontreal.ca>
parents:
diff changeset
108 super(SigmoidXEAutoEncoder, self)._instance_initialize(obj, input_size, hidden_size, **init)